Transfusion flow rate control optimization method based on improved PID (Proportion Integration Differentiation)

Abstract

The invention provides an infusion flow rate control optimization method based on improved PID (Proportion Integration Differentiation), which belongs to the field of PID control optimization and comprises the following specific steps: S1, acquiring key operation data in an infusion process by utilizing a real-time monitoring device; s2, dynamically matching the infusion flow rate change trend monitored in the infusion process with the response of the control system to generate an infusion flow rate error; s3, the infusion flow velocity error is input into an improved infusion flow velocity PID controller, and an infusion pump control signal is output; s4, according to the specific implementation of the improved infusion flow rate PID controller, the optimal control parameters of the infusion flow rate PID controller are obtained through iterative optimization of an improved simulated cooking training algorithm; the improved simulated cooking training algorithm comprises the following steps: S41, introducing a dynamic inertia weight; s42, using a global information transmission mechanism; s5, the infusion pump control signal is used for controlling the speed of an infusion pump, and accurate and stable control over the infusion flow speed is achieved.

Description

Technical Field

[0001] The invention belongs to the field of PID control optimization, and in particular relates to an infusion flow rate control optimization method based on improved PID. Background Art

[0002] The issue of infusion flow rate control is an important topic in the medical field, especially in the process of drug infusion therapy, accurate control of infusion rate is crucial to ensure patient safety; the infusion flow rate control system ensures that the patient receives the right amount of fluid, medicine and nutrition by adjusting the flow rate of the liquid in the infusion tube, thereby avoiding poor treatment effects or potential medical accidents caused by inaccurate flow rate; with the development of science and technology, especially the application of automation and intelligent control technology, precise control of infusion flow rate has become a standard requirement for modern hospital equipment; in addition, traditional manual control or simple mechanical control methods can no longer meet the high-precision requirements of modern medicine; with the development of microprocessors and sensor technology, intelligent and automated infusion flow rate control systems have gradually become mainstream; these systems can monitor the liquid flow rate and infusion volume in real time by integrating sensors, control units and feedback mechanisms, and automatically adjust the flow rate according to the set value, greatly improving the safety and accuracy of the treatment process.

[0003] Infusion flow rate control is often implemented using a PID control algorithm, which ensures that the infusion pump outputs liquid medicine at a desired flow rate by detecting the flow rate of the liquid medicine output by the infusion pump in real time and adjusting the input of the infusion pump control signal. PID control is a control algorithm widely used in control systems. By adjusting the proportional, integral and differential parameters, it can cope with rapid response changes in operating conditions and achieve real-time control of the system. However, it has the disadvantages of complex parameter setting and poor stability when facing nonlinear complex systems.

[0004] A New Human-Based Metahurestic Optimization MethodBased on Mimicking Cooking Training (CBOA) is a novel heuristic optimization method inspired by the experience accumulation, skill application and creative adjustment in the cooking process. This algorithm takes the learning process of cooking training as the basis for the optimization search strategy, simulating how chefs gradually master skills, adjust the proportion of ingredients, and control cooking time and temperature in actual training to achieve the best cooking results. This algorithm has good optimization performance, but to a certain extent, it is prone to fall into local optimal solutions and convergence prematurely. Summary of the invention

[0005] The purpose of the present invention is to achieve high-precision and stable control of infusion flow rate through an infusion flow rate control optimization method based on improved PID; to improve the simulation cooking training algorithm by introducing dynamic inertia weight and using a global information transmission mechanism, to improve the global search ability and local development ability of the algorithm, to enhance the algorithm optimization performance, and to avoid falling into local optimum; to establish a dynamic characteristic model of the infusion system, to adjust the infusion pump speed in real time in combination with the flow rate error, to ensure that the flow rate accurately tracks the set value, to optimize the infusion flow rate PID controller parameters Kp, Ki, Kd using the improved simulation cooking training algorithm, to output accurate infusion pump control signals, to improve the system's anti-interference ability and dynamic response speed, and to achieve reliability and efficiency of infusion flow rate control.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts an infusion flow rate control optimization method based on improved PID, and the specific steps are as follows.

[0007] S1. Use the real-time monitoring device to obtain key operating data during the infusion process, which is used to analyze the changes in infusion flow rate and system operating status.

[0008] S2. Dynamically match the infusion flow rate change trend monitored during the infusion process with the response of the control system to generate the infusion flow rate error .

[0009] S3, the infusion flow rate error Input to the improved infusion flow rate PID controller and output the infusion pump control signal.

[0010] S4. The improved infusion flow rate PID controller is specifically implemented as follows: an infusion flow rate model is established by simulating the flow rate dynamic characteristics of the infusion system, and the improved simulation cooking training algorithm is integrated with the infusion flow rate PID controller. The optimal control parameters Kp, Ki, and Kd of the infusion flow rate PID controller are obtained through iterative optimization of the improved simulation cooking training algorithm; the improved simulation cooking training algorithm includes: S41. In the algorithm exploration stage, dynamic inertia weights are introduced by adding periodic changes and the maximum value of inertia weights. , improve the mathematical model of position update of individuals in the population; S42, in the algorithm development stage, by introducing the influence of the global optimal solution, a global information transmission mechanism is used to improve the mathematical model of position update of individuals in the population.

[0011] S5. The infusion pump control signal is used to control the infusion pump speed to achieve accurate and stable control of the infusion flow rate.

[0012] Preferably, the infusion flow rate error in S2 is The expression is: ; In the formula, is the expected infusion flow rate, is the actual infusion flow rate.

[0013] Preferably, in said S4, an infusion flow rate model is established by simulating the flow rate dynamic characteristics of the infusion system, and the infusion flow rate model is: (1); In the formula, is the infusion flow rate, r is the inner radius of the infusion tube, is the viscosity of the liquid medicine, L is the length of the infusion tube, is the length of the pipeline of the infusion regulating device, is the density of the liquid, is the acceleration due to gravity, is the current liquid level height, is the liquid level height at the infusion terminal, is the cross-sectional area of ​​the drug liquid carrier, is the volume flow rate per unit time, and t is the time.

[0014] Preferably, the optimization goal of infusion flow rate control is mainly to ensure the accuracy, safety and stability of the infusion process; first, the system needs to accurately control the flow rate to ensure that the infusion volume is consistent with the set value to prevent overdose or underdose of drugs. This goal is the core requirement of the system; in addition, the optimization goals also include minimizing control errors, reducing errors caused by changes in pipeline resistance and fluctuations in liquid viscosity, and maintaining flow rate stability; the system should have rapid response and adjustment capabilities, and be able to adjust the flow rate in a dynamic environment in a timely manner to respond to changes in patient needs, ensuring a continuous and stable infusion process; in terms of system stability, the optimization goal is to improve the system's resistance to external disturbances, prevent unstable factors from affecting infusion accuracy, and ensure that the flow rate remains stable even when the environment changes or the equipment is damaged.

[0015] Preferably, the improved cooking simulation training algorithm in S4, wherein the S41 algorithm exploration phase introduces dynamic inertia weight by adding periodic changes and the maximum value of inertia weight , dynamic inertia weight The expression is: (2); In the formula, is the dynamic inertia weight, is the maximum value of the dynamic inertia weight, is the minimum value of the dynamic inertia weight, and is the period adjustment factor, is the current number of iterations, T is the maximum number of iterations; The mathematical model for updating the individual positions of the population in the exploration phase of the improved simulated cooking training algorithm is: (3); In the formula, is the new position of the i-th chef mentor in the j-th dimension, is the current position of the i-th chef mentor in the j-th dimension, r is a random number between 0 and 1, is the position of the current best chef mentor, and I is a random number in the set {1,2}.

[0016] Preferably, dynamic inertia weights are introduced in the exploration phase of the basic cooking simulation training algorithm The advantage of is that it can enhance the global search capability of the algorithm and accelerate the convergence process; specifically, the dynamic inertia weight provides a larger search step in the initial stage of optimization, which helps the algorithm to conduct extensive exploration in the solution space, thereby avoiding falling into the local optimal solution too early; when the algorithm is in the exploration stage, the larger inertia weight makes the search process have a strong randomness, which can quickly jump to the potential excellent solution area and increase the possibility of finding the global optimal solution; with the increase of the number of iterations, the inertia weight gradually decreases, and the search step of the algorithm also decreases accordingly, which helps the algorithm to reduce the randomness of the search when approaching the optimal solution, ensuring that the local search can be carried out more accurately and finally approaching the global optimal solution; this strategy effectively balances the relationship between global search and local search by adjusting the "roughness" and "fineness" of the search, avoiding overly concentrated local exploration in the early stage, and at the same time accelerating convergence by reducing the inertia weight in the later stage, preventing the search process from being too random, thereby improving the convergence speed and accuracy of the algorithm.

[0017] Preferably, in the improved cooking simulation training algorithm in S4, in the S42 algorithm development stage, the influence of the global optimal solution is introduced, and a global information transmission mechanism is used to improve the position update mathematical model of the population individuals: (4); In the formula, is the new position of the i-th chef student in the j-th dimension, is the current position of the i-th chef mentor in the j-th dimension, r is a random number between 0 and 1, The chef tutor is randomly selected by the student, I is a random number in the set {1,2}, is the period adjustment factor, Position for the best current chef students.

[0018] Preferably, the influence of the global optimal solution is introduced in the development stage of the basic simulated cooking training algorithm, and a global information transmission mechanism is used to significantly improve the performance and optimization effect of the algorithm; the introduction of the global optimal solution and the use of the global information transmission mechanism can effectively improve the exploration ability and convergence speed of the algorithm; first, by introducing the global optimal solution , the algorithm can refer to the position and state of the global optimal solution in real time, which is very important for avoiding falling into the local optimal solution; in the early stage of the algorithm, the breadth of the exploration space is usually large, and by transmitting relevant information about the global optimal solution, the algorithm can be accelerated to find potential excellent solution areas, thereby improving the global search ability of the algorithm; the transmission of the global optimal solution enables individuals to more clearly understand the optimal area of ​​the entire search space during the search process, which helps to guide them to quickly advance towards the optimal solution; secondly, the global information transmission mechanism is reflected in The use of this mechanism can promote information sharing between individuals and avoid a single individual from falling into the limitations of local search; each individual not only relies on its own local information during the search process, but can also use the global optimal solution and the information of other individuals to improve the efficiency of the search; this information sharing mechanism enables the entire algorithm to search more efficiently, thereby improving the convergence speed and accuracy.

[0019] Preferably, in said S4, the optimal control parameters Kp, Ki, Kd of the infusion flow rate PID controller are obtained by iterative optimization through the improved simulated cooking training algorithm, and the specific steps are: Step 1, simulate the working condition of the infusion flow rate control system and design the transfer function to describe the dynamic characteristics of the system; Step 2, encode the parameters Kp, Ki, and Kd of the infusion flow rate PID controller as the solution of the search space of the improved simulated cooking training algorithm. As the algorithm iterates, the updated individual positions of the population reflect the parameters of the corresponding infusion flow rate PID controller; Step 3, initialize the parameters of the improved simulated cooking training algorithm, including the algorithm's population size SearchAgents, problem dimension dimension, maximum number of iterations T, upper limit ub of the search space, and lower limit lb of the search space; Step 4: Calculate the fitness value of each individual in the current population according to the fitness function, compare the optimal fitness value of this iteration with the historical optimal value, and retain the optimal solution for subsequent iterations; Step 5, simulate and improve the optimization process of each stage of the simulated cooking training algorithm, that is, update the Kp, Ki, and Kd parameters of the infusion flow rate PID controller; Step 6. Determine whether the current number of iterations has reached the maximum number of iterations. If so, exit the loop, output the global optimal solution of the algorithm search space, and assign it to the infusion flow rate PID controller as the three parameters Kp, Ki, and Kd. Otherwise, return to execute Step 1.

[0020] Preferably, the Step 1 simulates the working condition of the infusion flow rate control system and designs a transfer function to describe the dynamic characteristics of the system. The transfer function formula is: ; ; ; ; In the formula, s represents the complex frequency domain variable in Laplace transform, K is the system gain, which reflects the influence of the control signal on the system output, C represents the discharge capacity of the infusion pump, is the total resistance of the system, including the pipeline resistance and the internal resistance of the infusion pump, and Respectively reflect the dynamic response characteristics of the infusion pump and the dynamic response characteristics of the infusion pipeline and fluid, represents the inertia of the infusion pump, is the viscous damping coefficient of the infusion pump, which indicates the influence of the internal friction and liquid resistance of the pump on the dynamic response, r is the inner radius of the infusion tube, L is the length of the infusion tube, is the density of the drug solution.

[0021] Preferably, in Step 2, the parameters Kp, Ki, and Kd of the infusion flow rate PID controller are encoded as the solution of the improved simulated cooking training algorithm search space. As the algorithm is iterated, the positions of the individuals in the population are updated, that is, the solution of the algorithm search space is updated, that is, the parameters of the infusion flow rate PID controller are updated. The encoding vector is: ; Where x is the encoding vector, i.e., the solution of the search space of the improved simulated cooking training algorithm, and Kp, Ki, and Kd are the proportional parameter, integral parameter, and differential parameter of the infusion flow rate PID controller, respectively.

[0022] Preferably, the Step 4 calculates the fitness value of each individual in the current population according to the fitness function, and the fitness function formula is: ; In the formula, is the fitness function, is the objective function value, λ is the regularization coefficient, is the regularization term.

[0023] Preferably, the Step 5 simulates and improves the optimization process of each stage of the simulated cooking training algorithm, and the specific steps are: Step 1: Simulate the algorithm exploration phase, introduce dynamic inertia weight by adding periodic changes and the maximum value of inertia weight , improve the mathematical model of position update of population individuals and implement formulas (2)-(3); Step 2: During the simulation algorithm development phase, by introducing the influence of the global optimal solution and using a global information transmission mechanism, the mathematical model of position update of individuals in the population is improved and formula (4) is implemented.

[0024] Preferably, the S5 infusion pump control signal mathematical model is: (5); In the formula, is the infusion pump control signal, is the weight factor, Kp, Ki, Kd are the three parameters of the infusion flow rate PID controller, is the infusion flow rate error, is the integral term of the infusion flow rate error, is the differential term of the infusion flow rate error, To globally optimize the control signal, as a reference for control signal correction, is the initial control signal, It is a disturbance compensation signal, which is used to correct the influence of external disturbance on the control signal. is the global information fusion coefficient, which weighs the influence of the initial control signal and the global optimization signal, and t is the time.

[0025] By adopting the above technical scheme, the beneficial effects of the present invention are as follows: by introducing dynamic inertia weights and using a global information transmission mechanism, the simulated cooking training algorithm is improved, the global search capability and local development capability of the algorithm are significantly improved, and the algorithm is prevented from falling into local optimality; the improved simulated cooking training algorithm is combined with the infusion flow rate PID controller to form an improved infusion flow rate PID controller, which accurately outputs the infusion pump control signal and improves the control capability of the infusion pump speed; at the same time, by comprehensively considering the flow rate error, adjustment time and the stability of the control signal, the infusion flow rate is stably controlled, the rapid response of the system and the dynamic performance optimization are guaranteed, and a safe, reliable and effective solution is provided for high-precision infusion flow rate control in medical scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flow chart of an infusion flow rate control optimization method based on improved PID.

[0027] Figure 2Flowchart of steps for iteratively optimizing the best control parameters of the infusion flow rate PID controller to improve the simulated cooking training algorithm.

[0028] Figure 3 A fitness value comparison curve chart of the basic simulated cooking training algorithm and the improved simulated cooking training algorithm.

[0029] Figure 4 Process diagram of optimizing parameter value Kp for basic simulated cooking training algorithm and improved simulated cooking training algorithm.

[0030] Figure 5 Process diagram of optimizing parameter value Ki for basic simulated cooking training algorithm and improved simulated cooking training algorithm.

[0031] Figure 6 Process diagram of optimizing parameter value Kd for basic simulated cooking training algorithm and improved simulated cooking training algorithm.

[0032] Figure 7 A comparison chart of the PID control effect optimization between the basic simulation cooking training algorithm and the improved simulation cooking training algorithm. DETAILED DESCRIPTION

[0033] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0034] The present invention provides a technical solution: an infusion flow rate control optimization method based on improved PID, which specifically includes the following steps: Figure 1 shown.

[0035] S1. Use the real-time monitoring device to obtain key operating data during the infusion process, which is used to analyze the changes in infusion flow rate and system operating status.

[0036] S2. Dynamically match the infusion flow rate change trend monitored during the infusion process with the response of the control system to generate the infusion flow rate error .

[0037] Furthermore, the infusion flow rate error in S2 The expression is: ; In the formula, is the expected infusion flow rate, is the actual infusion flow rate.

[0038] S3, the infusion flow rate error Input to the improved infusion flow rate PID controller and output the infusion pump control signal.

[0039] S4. The improved infusion flow rate PID controller is specifically implemented as follows: an infusion flow rate model is established by simulating the flow rate dynamic characteristics of the infusion system, and the improved simulation cooking training algorithm is integrated with the infusion flow rate PID controller. The optimal control parameters Kp, Ki, and Kd of the infusion flow rate PID controller are obtained through iterative optimization of the improved simulation cooking training algorithm; the improved simulation cooking training algorithm includes: S41. In the algorithm exploration stage, dynamic inertia weights are introduced by adding periodic changes and the maximum value of inertia weights. , improve the mathematical model of position update of individuals in the population; S42, in the algorithm development stage, by introducing the influence of the global optimal solution, a global information transmission mechanism is used to improve the mathematical model of position update of individuals in the population.

[0040] Furthermore, in S4, an infusion flow rate model is established by simulating the flow rate dynamic characteristics of the infusion system. The infusion flow rate model is: (1); In the formula, is the infusion flow rate, r is the inner radius of the infusion tube, is the viscosity of the liquid medicine, L is the length of the infusion tube, is the length of the pipeline of the infusion regulating device, is the density of the liquid, is the acceleration due to gravity, is the current liquid level height, is the liquid level height at the infusion terminal, is the cross-sectional area of ​​the drug liquid carrier, is the volume flow rate per unit time, and t is the time.

[0041] Furthermore, the improved cooking simulation training algorithm in S4, wherein the S41 algorithm exploration phase introduces a dynamic inertia weight by adding periodic changes and the maximum value of the inertia weight. , dynamic inertia weight The expression is: (2); In the formula, is the dynamic inertia weight, is the maximum value of the dynamic inertia weight, is the minimum value of the dynamic inertia weight, and is the period adjustment factor, is the current number of iterations, T is the maximum number of iterations; The mathematical model for updating the individual positions of the population in the exploration phase of the improved simulated cooking training algorithm is: (3); In the formula, is the new position of the i-th chef mentor in the j-th dimension, is the current position of the i-th chef mentor in the j-th dimension, r is a random number between 0 and 1, is the position of the current best chef mentor, and I is a random number in the set {1,2}.

[0042] Furthermore, the improved cooking simulation training algorithm in S4, wherein the S42 algorithm development stage introduces the influence of the global optimal solution and uses a global information transmission mechanism to improve the position update mathematical model of the population individuals as follows: (4); In the formula, is the new position of the i-th chef student in the j-th dimension, is the current position of the i-th chef mentor in the j-th dimension, r is a random number between 0 and 1, The chef tutor is randomly selected by the student, I is a random number in the set {1,2}, is the period adjustment factor, Position for the best current chef students.

[0043] Furthermore, in S4, the optimal control parameters Kp, Ki, and Kd of the infusion flow rate PID controller are obtained by iterative optimization through the improved simulated cooking training algorithm, such as Figure 2 As shown, the specific steps are: Step 1, simulate the working condition of the infusion flow rate control system and design the transfer function to describe the dynamic characteristics of the system; Step 2, encode the parameters Kp, Ki, and Kd of the infusion flow rate PID controller as the solution of the search space of the improved simulated cooking training algorithm. As the algorithm iterates, the updated individual positions of the population reflect the parameters of the corresponding infusion flow rate PID controller; Step 3, initialize the parameters of the improved simulated cooking training algorithm, including the algorithm's population size SearchAgents, problem dimension dimension, maximum number of iterations T, upper limit ub of the search space, and lower limit lb of the search space; Step 4: Calculate the fitness value of each individual in the current population according to the fitness function, compare the optimal fitness value of this iteration with the historical optimal value, and retain the optimal solution for subsequent iterations; Step 5, simulate and improve the optimization process of each stage of the simulated cooking training algorithm, that is, update the Kp, Ki, and Kd parameters of the infusion flow rate PID controller; Step 6. Determine whether the current number of iterations has reached the maximum number of iterations. If so, exit the loop, output the global optimal solution of the algorithm search space, and assign it to the infusion flow rate PID controller as the three parameters Kp, Ki, and Kd. Otherwise, return to execute Step 1.

[0044] Furthermore, the Step 1 simulates the working condition of the infusion flow rate control system and designs a transfer function to describe the dynamic characteristics of the system. The transfer function formula is: ; ; ; ; In the formula, s represents the complex frequency domain variable in Laplace transform, K is the system gain, which reflects the influence of the control signal on the system output, C represents the discharge capacity of the infusion pump, is the total resistance of the system, including the pipeline resistance and the internal resistance of the infusion pump, and Respectively reflect the dynamic response characteristics of the infusion pump and the dynamic response characteristics of the infusion pipeline and fluid, represents the inertia of the infusion pump, is the viscous damping coefficient of the infusion pump, which indicates the influence of the internal friction and liquid resistance of the pump on the dynamic response, r is the inner radius of the infusion tube, L is the length of the infusion tube, is the density of the drug solution.

[0045] Furthermore, the Step 2 encodes the parameters Kp, Ki, and Kd of the infusion flow rate PID controller into the solution of the improved simulated cooking training algorithm search space. As the algorithm is iterated, the individual positions of the population are updated, that is, the solution of the algorithm search space is updated, that is, the parameters of the infusion flow rate PID controller are updated. The encoding vector is: ; Where x is the encoding vector, i.e., the solution of the search space of the improved simulated cooking training algorithm, and Kp, Ki, and Kd are the proportional parameter, integral parameter, and differential parameter of the infusion flow rate PID controller, respectively.

[0046] Furthermore, the Step 4 calculates the fitness value of each individual in the current population according to the fitness function, and the fitness function formula is: ; In the formula, is the fitness function, is the objective function value, λ is the regularization coefficient, is the regularization term.

[0047] Furthermore, the Step 5 simulates and improves the optimization process of each stage of the simulated cooking training algorithm, and the specific steps are: Step 1: Simulation algorithm exploration phase, by adding periodic changes and the maximum value of inertia weight, introduce dynamic inertia weight , improve the mathematical model of position update of population individuals and implement formulas (2)-(3); Step 2: During the simulation algorithm development phase, by introducing the influence of the global optimal solution and using a global information transmission mechanism, the mathematical model of position update of individuals in the population is improved and formula (4) is implemented.

[0048] S5. The infusion pump control signal is used to control the infusion pump speed to achieve accurate and stable control of the infusion flow rate.

[0049] Furthermore, the mathematical model of the S5 infusion pump control signal is: (5); In the formula, is the infusion pump control signal, is the weight factor, Kp, Ki, Kd are the three parameters of the infusion flow rate PID controller, is the infusion flow rate error, is the integral term of the infusion flow rate error, is the differential term of the infusion flow rate error, To globally optimize the control signal, as a reference for control signal correction, is the initial control signal, It is a disturbance compensation signal, which is used to correct the influence of external disturbance on the control signal. is the global information fusion coefficient, which weighs the influence of the initial control signal and the global optimization signal, and t is the time.

[0050] In order to verify the infusion flow rate control optimization method based on improved PID proposed in the present invention, simulation experiments were carried out through Matlab and Simulink, and the PID control optimization of the improved simulation cooking training algorithm and the basic simulation cooking training algorithm was compared. The mathematical model of the basic simulation cooking training algorithm was first improved through Matlab, and the PID control model of the infusion flow rate control system was established through Simulink; the parameters of the improved simulation cooking training algorithm were initialized, including the algorithm population size SearchAgents=30, the problem dimension dimension=3, the maximum number of iterations T=30, the upper limit ub=1 of the search space, and the lower limit lb=0 of the search space. The Matlab program was run to obtain a fitness value comparison curve chart of the basic simulation cooking training algorithm and the improved simulation cooking training algorithm, as shown in FIG. Figure 3As shown in the figure, the improved simulated cooking training algorithm reached the optimal fitness value of 27.92 in the 19th iteration, and it can be seen from the fitness curve that the improved algorithm has significant advantages over the basic algorithm; first, the improved algorithm shows a faster convergence speed in the optimization process, and the fitness value drops rapidly in the first few generations, indicating that its global search ability has been enhanced, and it can explore the solution space more efficiently and find potential excellent solution areas in the early stage; while the fitness value of the basic algorithm drops more slowly, indicating that its initial search efficiency is low; secondly, the final fitness value of the improved algorithm is significantly lower than that of the basic algorithm, reflecting a higher optimization accuracy, which indicates that the improved algorithm has a stronger local development ability in the later stage of optimization, and can further refine the search and find a better solution; Overall, the improved algorithm not only speeds up the convergence speed of the optimization process, but also significantly improves the final solution quality and search efficiency, reflecting the balanced optimization of global search and local search, indicating that the improved algorithm can more effectively optimize the performance of the infusion flow rate control system, thereby achieving a more accurate and stable infusion flow rate control effect.

[0051] Furthermore, if Figure 4-6 As shown, the improved simulated cooking training algorithm is used to adjust the parameters of the flow rate PID controller of the infusion flow rate control system. The optimal PID control parameters obtained by the improved algorithm are Kp=0.15, Ki=0.13, Kd=0.01, and the optimal PID control parameters obtained by the basic algorithm are Kp=0.16, Ki=0.15, Kd=0.13; the optimal control parameters are input into the infusion flow rate PID controller to obtain the optimal control effect of the infusion flow rate control system; Figure 7 A comparison chart of the PID control effects optimized by the basic simulation cooking training algorithm and the improved simulation cooking training algorithm is shown, and the set value is set to 3 units. It can be seen from the figure that the improved simulation cooking training algorithm shows better performance than the basic algorithm in optimizing the PID controller; first, in terms of the response speed of the system, the PID controller optimized by the improved algorithm can approach the set value faster, indicating that its dynamic response is faster, the adjustment ability in the initial stage is stronger, and the infusion flow rate can quickly approach the set value; secondly, the overshoot is significantly reduced during the adjustment process, and the infusion process is smoother, avoiding excessive drug or flow rate fluctuations caused by overshoot, and ensuring the safety of infusion; in addition, the improved controller can enter the steady state faster, reflecting the high-precision flow rate control capability, ensuring the accuracy and stability of the infusion dosage; this optimization not only improves the accuracy and safety of the infusion control system, but also reduces the influence of dynamic errors on the control effect, ensuring a more efficient and reliable flow rate control effect for the infusion scenario.

Claims

1. An infusion flow rate control optimization method based on improved PID, characterized in that: The specific steps are: S1. Use the real-time monitoring device to obtain key operating data during the infusion process, which is used to analyze the changes in infusion flow rate and system operating status; S2. Dynamically match the infusion flow rate change trend monitored during the infusion process with the response of the control system to generate the infusion flow rate error ; S3, the infusion flow rate error Input to the improved infusion flow rate PID controller, output infusion pump control signal; S4. The improved infusion flow rate PID controller is specifically implemented as follows: an infusion flow rate model is established by simulating the flow rate dynamic characteristics of the infusion system, an improved simulated cooking training algorithm is integrated with the infusion flow rate PID controller, and the optimal control parameters Kp, Ki, and Kd of the infusion flow rate PID controller are obtained through iterative optimization of the improved simulated cooking training algorithm; The improved cooking simulation training algorithm includes: S41, in the algorithm exploration phase, a dynamic inertia weight is introduced by adding periodic changes and the maximum value of the inertia weight. , improve the mathematical model of position update of individuals in the population; S42, in the algorithm development stage, by introducing the influence of the global optimal solution, a global information transmission mechanism is used to improve the mathematical model of position update of individuals in the population; S5. The infusion pump control signal is used to control the infusion pump speed to achieve accurate and stable control of the infusion flow rate.

2. The infusion flow rate control optimization method based on improved PID according to claim 1, characterized in that: In S4, an infusion flow rate model is established by simulating the flow rate dynamic characteristics of the infusion system. The infusion flow rate model is: （1）； In the formula, is the infusion flow rate, r is the inner radius of the infusion tube, is the viscosity of the liquid medicine, L is the length of the infusion tube, is the length of the pipeline of the infusion regulating device, is the density of the drug solution, is the acceleration due to gravity, is the current liquid level height, is the liquid level height at the infusion terminal, is the cross-sectional area of ​​the drug liquid carrier, is the volume flow rate per unit time, and t is the time.

3. The infusion flow rate control optimization method based on improved PID according to claim 2, characterized in that: The improved cooking simulation training algorithm in S4, wherein the S41 algorithm exploration phase introduces dynamic inertia weight by adding periodic changes and the maximum value of inertia weight , dynamic inertia weight The expression is: （2）； In the formula, is the dynamic inertia weight, is the maximum value of the dynamic inertia weight, is the minimum value of the dynamic inertia weight, and is the period adjustment factor, is the current number of iterations, T is the maximum number of iterations; The mathematical model for updating the individual positions of the population in the exploration phase of the improved simulated cooking training algorithm is: （3）； In the formula, is the new position of the i-th chef mentor in the j-th dimension, is the current position of the i-th chef mentor in the j-th dimension, r is a random number between 0 and 1, is the position of the current best chef mentor, and I is a random number in the set {1,2}.

4. The infusion flow rate control optimization method based on improved PID according to claim 3, characterized in that: The improved cooking simulation training algorithm in S4, wherein the S42 algorithm development stage introduces the influence of the global optimal solution and uses a global information transmission mechanism to improve the position update mathematical model of the population individuals as follows: （4）； In the formula, is the new position of the i-th chef student in the j-th dimension, is the current position of the i-th chef mentor in the j-th dimension, r is a random number between 0 and 1, The chef tutor is randomly selected by the student, I is a random number in the set {1,2}, is the period adjustment factor, Position for the best current chef students.

5. The infusion flow rate control optimization method based on improved PID according to claim 4, characterized in that: In S4, the optimal control parameters Kp, Ki, and Kd of the infusion flow rate PID controller are obtained by iterative optimization through the improved simulated cooking training algorithm. The specific steps are: Step 1, simulate the working condition of the infusion flow rate control system and design the transfer function to describe the dynamic characteristics of the system; Step 2, encode the parameters Kp, Ki, and Kd of the infusion flow rate PID controller as the solution of the search space of the improved simulated cooking training algorithm. As the algorithm iterates, the updated individual positions of the population reflect the parameters of the corresponding infusion flow rate PID controller; Step 3, initialize the parameters of the improved simulated cooking training algorithm, including the algorithm's population size SearchAgents, problem dimension dimension, maximum number of iterations T, upper limit ub of the search space, and lower limit lb of the search space; Step 4: Calculate the fitness value of each individual in the current population according to the fitness function, compare the optimal fitness value of this iteration with the historical optimal value, and retain the optimal solution for subsequent iterations; Step 5, simulate and improve the optimization process of each stage of the simulated cooking training algorithm, that is, update the Kp, Ki, and Kd parameters of the infusion flow rate PID controller; Step 6. Determine whether the current number of iterations has reached the maximum number of iterations. If so, exit the loop, output the global optimal solution of the algorithm search space, and assign it to the infusion flow rate PID controller as the three parameters Kp, Ki, and Kd. Otherwise, return to execute Step 1.

6. The infusion flow rate control optimization method based on improved PID according to claim 5, characterized in that: The mathematical model of the S5 infusion pump control signal is: （5）； In the formula, is the infusion pump control signal, is the weight factor, Kp, Ki, Kd are the three parameters of the infusion flow rate PID controller, is the infusion flow rate error, is the integral term of the infusion flow rate error, is the differential term of the infusion flow rate error, To globally optimize the control signal, as a reference for control signal correction, is the initial control signal, It is a disturbance compensation signal, which is used to correct the influence of external disturbance on the control signal. is the global information fusion coefficient, which weighs the influence of the initial control signal and the global optimization signal, and t is the time.

Images